A roll-over protective structure (ROPS), as described in ASAE S519, is a protective structure designed to minimize the frequency and severity of operator injury resulting from accidental tractor upset (ASAE, 1996). ROPS utilization has shown to have dramatic impacts in reducing fatalities during tractor overturns. But some agricultural tractors do not have appropriate ROPS designs and ROPS are not available. This project investigates and evaluates roll-over protective structure (ROPS) designs for agricultural tractors in the United States to provide operator protection on tractors and in operating conditions not currently available. Initially, an inventory of agricultural tractor ROPS availability (specifically for older tractors) was conducted. Classification of axle housings and determination of the population of tractors for which ROPS are not available was performed. The most popular pre-ROPS tractors were categorized into 4 major axle categories, with 4 tractors (Ford 8N, John Deere A, Farmall M and Allis Chalmers D 17), representing each major axle category. As appropriate ROPS designs had previously been conducted for the Ford 8N and Farmall M, ROPS were designed, constructed and tested, according to ASAE S519 (SAE J2194) for the John Deere A and Allis Chalmers D 17 tractors. Both static and field upset tests for side and rear rollover were conducted in accordance with ASAE S519 (SAE 12194). Measured deflection at the energy criteria and during field upset tests were less than the allowable deflection, indicating successful Raps designs. Radio-controlled John Deere A and Allis Chalmers D 17 tractors were utilized in the field upset tests. The axle housing strength tests were conducted on the John Deere A, Allis Chalmers D 17 and Allis Chalmers WD 45 tractor models to evaluate the suitability of the axle housing to support a Raps. All axle-housing strengths exceeded the stress applied during the static longitudinal tests, with a factor of safety greater than 1.6. This report also presents the results of dynamic deflection of ROPS when John Deere A tractor equipped with radio controlled system was subjected to overturn and analyzes the effect of impulse loading on Raps design and installation pattern. In order to measure the dynamic deflection of ROPS, a L VDT (Linear Variable Differential Transformer) was installed in same plane in which the deflection of ROPS occurred. The position of tractor was monitored by the MSTS (measuring system of tractor stability). Tractor position, ROPS rotation velocities and dynamic deflection of Raps were recorded at a 40 Hz acquisition rate. This information can be useful for deployment times for the NIOSH auto-ROPS. The time from instability to the ROPS striking the ground ranged from 0.7 to 1.0 seconds during the field-upset tests. Drawing of the ROPS designs for the John Deere A and Allis Chalmers D 17 tractors are presented. This project demonstrates the procedure and feasibility for designing and testing ROPS for pre-ROPS tractors.